Repeatability, Optimal Sample Size of Measurement and Phenotypic Correlations of Quantitative Traits in Guava

Abstract

Five fruits from each of 11 guava genotypes were evaluated in dry and early rainy seasons under Thailand conditions for fruit weight, flesh thickness, flesh weight, seed cavity (central pulp) weight, fruit firmness, total soluble solids, titratable acidity, juice acidity, and ascorbic acid to estimate repeatability (R), phenotypic correlations (r), and to predict the optimal sample size. The repeatability of the fruit weight, flesh thickness, flesh weight, seed cavity weight, titratable acidity, juice acidity, and ascorbic acid were relatively high (R ≥ 0.60). The flesh thickness, titratable acidity, juice acidity, and ascorbic acid were the traits with highest estimates, 0.85, 0.85, 0.87, 0.76 and 0.85, 0.83, 0.84, 0.80 in dry and early rainy seasons, respectively. Based on a threshold of 10% increase in relative efficiency, a sample of three fruits was sufficient for evaluating guava fruit traits. Most physical traits (fruit weight, flesh thickness, flesh weight, and seed cavity weight) had weak negative correlations (-0.25 ≤ r ≤ –0.38) with chemical traits (total soluble solids, titratable acidity, and ascorbic acid). Fruit firmness had no correlation with all other fruit traits. There were strong positive correlations between fruit weight and flesh thickness (r = 0.81), flesh weight (r = 0.99), and seed cavity weight (r = 0.88). Therefore, fruit weight could be used as an indirect selection for flesh thickness, flesh weight, and seed cavity weight.